Character And Pattern Recognition TRIZ Cases
- Vehicle Monitoring Camera Switching Imaging Conditions for Feature Identification
A vehicle monitoring camera switches imaging conditions to capture distinct features like license plates and occupant faces.
- Task Exchange Server for Human Media Review Coordination
A task exchange server coordinates human reviewers to identify targets in media data through distributed task assignment.
- Low Rank Sparsity Model for Video Anomaly Detection
A low rank sparsity prior model reconstructs event encodings from video footage to identify anomalies.
- End-to-End Modeling With Code Enhancement Layer
Target-based end-to-end model uses code enhancement layer to add target unit information to feature sequences.
- Graphical Risk Model for Unbiased Data Collection
Gesture-driven graphical models replace biased surveys to visualize interconnected risks.
- Text Wrap Detection Using Image Segmentation and Machine Learning
A text wrap module segregates digital content into groups and overlays graphical elements to generate feature maps for machine learning analysis.
- Latent Variable Generative Model Noise Contrastive Prior
Classifier reweights prior samples to match encoder distribution, resolving the prior hole problem in variational autoencoders.
- Neural Network Self-Supervised Labeling via Image Perturbation
Neural network generates object class probabilities and applies image perturbations to reduce manual labeling costs for restricted images.
- Deep Image Model Translation for Synthetic Data Augmentation
A deep image model translation system generates realistic training images from synthetic sources using convolutional autoencoders and generative adversarial networks.
- CD Metrology System Classifying Semiconductor Structural Elements
A CD metrology system classifies similar structural elements using stage-specific feature assessment from semiconductor images.
- Movable Imaging Terminal for True Stereo Dimension Measurement
A movable imaging terminal adjusts its optical axis angle to capture stereo images for precise object dimension calculation.
- 3D Imaging System Using Event Cameras for Mixed Reflectance Surfaces
A 3D imaging system uses event cameras and a projector to capture scene data for precise reconstruction.
- Vehicle Lane Localization via Sensor Fusion and HD Map Alignment
A lane level localization system fuses perception sensors with high-definition map data to align detected road features for precise vehicle positioning.
- 3D Medical Imaging Analysis with Interconnected Visual Outputs
Processor generates simultaneous overview, cross-sectional, and statistical images for three-dimensional medical data analysis.
- Document Image Fingerprinting for Predictive Structural Analysis
Multi-dimensional fingerprint analysis resolves the trade-off between computational efficiency and measurement precision in predictive structural analysis.
- Associative Pattern Memory Cycle Detection for High-Speed Recall
Associative pattern memory uses vertical sensors and cycle detection algorithms to enable high-speed fuzzy recognition of input patterns.
- Hand Gesture Recognition Using HOG Descriptors and Skin Color Filtering
A hand gesture recognition method uses HOG descriptors to identify dynamic organ movements within scanned image regions.
- Automated Driving Complexity Metric and Resource Allocation
A vehicle control system computes an environmental complexity metric to dynamically adjust operational modes.
- Video Label Augmentation via Bidirectional Propagation
Bidirectional label propagation augments video data to resolve the trade-off between labeling accuracy and time consumption.
- CNN Operation Device Skipping Zero Pixels via Mosaic Prediction
A convolutional neural network operation device predicts zero outputs to skip unnecessary pixel calculations.
- Automated Report Generation System Using Machine Learning
A system generates professional reports using machine learning models and sentiment analysis to suggest content from structured libraries.
- Computer Vision Flare Stack Emission Monitoring
Automated image analysis replaces manual visual inspections of flare stacks, reducing inspection time and human error.
- Real-Time Video Analysis for Dynamic Individual Identification
Intermediary processing segments complex analysis tasks, enabling accurate identification of specific individuals without increasing mobile device complexity.
- Joint Color and Depth Encoding for 3D Video Compression
Segmenting depth data via color-based confidence values improves compression efficiency while maintaining high-quality 3D video reconstruction.
- Augmented Reality Gaming Overlay for Mobile Devices
Mobile devices overlay virtual 3D components onto physical gaming machines to enhance player engagement without increasing hardware complexity.
- IF-Net Image Processing for Lighting Invariant Feature Matching
Dynamic IF-Net training with light interference loss improves matching proficiency under varying outdoor lighting conditions.
- Projective Distortion Correction Using Eigenpoints
A projective correction method using eigenpoints to determine vanishing points for image text baselines.
- Random Pigment Domain Marking for High-Throughput Object Identification
Extracting coordinates from random pigment domains reduces IT storage while maintaining high-throughput object identification.
- Vehicle Lane Determination Using Map Reference and Sensor Data
A vehicle controller compares reference lane line information with estimated sensor data to determine the driving lane.
- Vehicle Vision System Lane-Specific Speed Limit Validation
A vehicle vision system processes camera imagery to identify traffic signs and determine applicable speed limits for the current travel lane.
- Cotton Feature Identification Using Stitched Point Clouds
Stitched point clouds identify cotton plant features using depth data, replacing manual inspection to improve assessment throughput.
- Video Panoptic Segmentation via Temporal Feature Fusion
A video panoptic segmentation method fuses target and reference frame features using a spatial-temporal attention module to generate consistent object masks.
- 3D Camera Road Friction Estimation via Height Profile Analysis
A 3D camera creates road height profiles to estimate friction coefficients using surface feature detection.
- AI Roof Detection for Automated Estimation Reports
Automated roof detection system replaces manual site visits with aerial image analysis to generate accurate estimation reports.
- Pixel Data Compression Device for Low Gray Level Quality
A pixel data compression device adapts compressed data formats based on bit comparisons to preserve low gray level precision.
- Imaging Flame Detection with Zonal Criticality Weighting
An imaging system assigns pixel arrays to zones of varying criticality levels for targeted flame status determination.
- Cognitive Function Estimation Using Multi-Source Driver Data
A cognitive function estimation device processes vehicle, face, and biological information to assess driver state.
- Automated Architecture Diagram Integration via Image Feature Mapping
Image feature mapping identifies touchpoints across separate architecture diagrams to generate a unified view, reducing manual analysis effort.
- Distributed Dataset Distillation for Privacy-Preserving Edge Model Training
Dataset distillation process compresses edge data streams into condensed representations for pre-training machine learning models.
- Synthetic Barcode Module Emulating Reflected Light for Secure Scanning
A synthetic barcode module emits light pulses to emulate reflected light from a scanned barcode.
- Image Sequence Detection Using Historical Feature Fusion
A pretrained target detection model incorporates historical data features alongside current frames to improve image sequence classification accuracy.
- Safety Assessment Apparatus Using Multi-Source Sensor Fusion
Information processing apparatus detects persons and objects to recognize actions based on their relevance.
- Semantic Segmentation Model Training via Self-Supervised Sub-Image Extraction
A semantic segmentation model extracts sub-images from unlabeled data to train neural networks without manual labeling.
- Biometric Image Sorting via Face Template Matching
Face detection systems extract biometric templates to associate images, resolving number obscuration issues during large-scale events.
- Voice Input Classification Using Visual Audio Data
A voice-control system classifies second inputs as follow-on requests or deliberation using visual and audio data.
- Dual-Imager Self-Checkout Tracking via Steering Mirror
A dual-imager system uses a steering mirror to direct a scanning imager's field of view toward tracked objects for high-resolution capture.
- Barcode Selection via Context Filtering in Camera Systems
Camera system filters detected barcodes by device context to display relevant payment links, reducing user selection time and confusion.
- Face Detection Method Using Dual-Process Segmentation
A face detection method segments input images by inclination to apply distinct processing routines for each category.
- Automated Plant Monitoring via Machine Vision and Predictive Modeling
Machine vision captures plant images for automated condition analysis using trained prediction models.